vLLM Semantic Router has announced the release of d3-ultra, a cutting-edge multimodal decision model with 27 billion parameters. This innovative model is designed to read text, images, and video in a single pass, significantly enhancing its efficiency and performance.
The d3-ultra model returns probabilities for various answers, demonstrating its versatility across different data types. It has achieved an impressive score of 87.2 on the team’s SR-Text test, showcasing its capabilities in understanding and processing complex information.
This release marks a significant advancement in the field of AI, particularly in multimodal applications, as it allows for more integrated and comprehensive decision-making processes. The open-weight nature of the model also encourages further research and development in the AI community.
Compiled automatically by the Tech AI Newsdesk from public AI-news sources and summarised in our own words.
